Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 23266
    Prev
    Next
    INDEX
    Explanations

    phrases indicating comparison or similarity

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    430
    +0.09
    0.4%
    494
    +0.09
    0.4%
    676
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    512
    +0.09
    0.02
    676
    +0.09
    0.02
    494
    +0.09
    0.03
    Negative Logits
     Louie
    -0.72
     generally
    -0.64
     wisely
    -0.63
     preferably
    -0.59
     Shad
    -0.59
     typically
    -0.58
     reportedly
    -0.58
     principally
    -0.57
     purportedly
    -0.56
     Regions
    -0.56
    POSITIVE LOGITS
    rontal
    0.70
    itals
    0.68
    aneously
    0.67
    istic
    0.65
    haust
    0.64
     spoiler
    0.64
    bones
    0.63
    ortal
    0.61
    uphem
    0.61
    complete
    0.61
    Activations Density 0.118%

    No Known Activations